Journal of Avian Biology 46: 9–17, 2015

doi: 10.1111/jav.00496
© 2014 The Authors. Journal of Avian Biology © 2014 Nordic Society Oikos
Subject Editor: Alexandre Roulin. Editor-in-Chief: Jan-Åke Nilsson. Accepted 14 July 2014

Sexual pigmentation and parental risk-taking in yellow warblers

Setophaga petechia

Andrea S. Grunst, Melissa L. Grunst and John T. Rotenberry­

A. S. Grunst (agrun001@ucr.edu), Univ. of California, Riverside – Biology, 900 University Ave, Riverside, CA 92507, USA. – M. L. Grunst,
Univ. of California, Riverside – Biology, Biology Dept, Riverside, CA 92521, USA. – J. T. Rotenberry, Univ. of Minnesota – Biological Sciences,
St Paul, MN, USA.­

Adult-directed predation risk imposes important behavioral constraints on parents and might thus alter relationships
between costly sexual ornaments and parental performance. For instance, under low predation risk, highly ornamented
individuals might display better parental performance than others, as predicted by ‘good parent’ models of sexual selec-
tion. However, under high risk of predation, highly ornamented individuals might abandon parental effort if conspicuous
to predators, or if social partners are more willing to take parental risks when paired with highly ornamented mates.
We experimentally elevated perceived adult-directed predation risk near nests to explore how carotenoid- and phaeomel-
anin-based pigmentation in both sexes relate to parental risk-taking for offspring in the yellow warbler Setophaga petechia.
Compared to other males, males with more intense carotenoid-based pigmentation maintained higher levels of paternal
effort under predation risk at highly concealed nests, but reduced nestling provisioning rate more at exposed nests. Further,
when faced with predation risk, females with more phaeomelanin-based pigmentation reduced nestling provisioning rate
less than other females, regardless of nest concealment. Females displayed higher parental effort across treatments when
paired to males with more colorful carotenoid pigmentation. However, birds did not reduce parental effort under risk
less when paired to a highly ornamented mate, suggesting that predation risk did not accentuate differential allocation.
Males did not take fewer parental risks than females. Results indicate that nest concealment modifies parental risk-taking
by males with colorful carotenoid-based pigmentation, and suggest that female melanin-based pigmentation may indicate
boldness and greater a propensity to take parental risks.

Adult-directed predation risk imposes important behavioral
constraints on parents, and might thus alter relationships
between costly sexual ornaments and parental behaviors
in a variety of alternative ways. A deficit of studies renders
these alternatives and the behavioral processes involved
poorly understood. For instance, under low levels of pre-
dation risk, highly ornamented individuals might display
better parental performance than others, as predicted by
‘good parent’ models of sexual selection (Hoelzer 1989,
Hill 1991, Siefferman and Hill 2003). In contrast, under
high predation risk, highly ornamented individuals might
abandon parental effort if vulnerable to detection or cap-
ture by predators (Martin and Badyaev 1996, Götmark
and Olsson 1997, Cabido et  al. 2008). Alternatively,
highly ornamented individuals might be more willing
to maintain parental effort under predation risk if better
able to evade predators (Fowler-Finn and Hebets 2011), if
ornamentation indicates higher investment towards cur-
rent reproduction rather than survival (Candolin 1998),
or if ornamentation correlates with behavioral character-
istics such as boldness and aggressiveness (Roulin 2004,
Ducrest et al. 2008, Da Silva et al. 2013). In these latter cases,
ornamentation may signal a willingness to take parental
risks to potential mates. A link between boldness, aggres-
siveness and pigmentation has been particularly proposed
for melanin-based pigmentation (Ducrest et al. 2008). For
instance, male great tits Parus major with large eumelanin-
based (black) plumage badges defend nests against predators
more vigorously than other males (Quesada and Senar 2007),
and female tawny owls Strix aluco with darker phaeomel-
anin-based (a reddish, cysteine-bearing form of melanin)
pigmentation are more aggressive towards threats to the nest
(Da Silva et  al. 2013). However, other types of coloration
have also been positively associated with boldness in the face
of a predation threat, although in non-parental contexts.
For instance, male guppies Poecilia reticulata with colorful
carotenoid-based coloration are more likely to approach
cichlid predators, potentially because these individuals are
better able to escape predators (Godin and Dugatkin 1996).
In addition, in biparental species, predation risk might
also affect the discrepancy in parental investment between
individuals with ‘attractive’ and ‘unattractive’ mates, and
between members of a mated pair (Burley 1986, Matessi
et al. 2009). The differential allocation hypothesis proposes

9
that individuals paired to elaborately ornamented mates
invest more into parental effort due to high brood value,
which might allow their mate to down-regulate effort
(Burley 1986, Møller and Thornhill 1998, Harris and Uller
2009, Ratikainen and Kokko 2010). For instance, male rock
sparrows Petronia petronia paired to females with enhanced
sexual pigmentation engaged in higher levels of risky nest
defense behavior against a weasel Mustela nivalis predator
(Matessi et al. 2009). However, males paired to females with
enhanced ornaments did not contribute more to lower-risk
nestling provisioning behavior (Matessi et  al. 2009), sug-
gesting that higher costs of parental care might accentuate
differential allocation. On the other hand, other studies
have found the opposite pattern. For example, male blue tits
Cyanistes caeruleus paired to females with reduced ultraviolet
reflectance invested less in nestling provisioning, but did
not defend nestlings against an aesculapian snake Zamenis
longissimus at reduced intensity (Mahr et al. 2012).
Birds provide good study species for exploring how the
sexual ornamentation of focal individuals and social mates
relates to levels of parental risk-taking for offspring, since
biparental care is prevalent and effects of adult-directed
predation risk on parental behaviors are well documented.
High rates of visitation to a stationary nest render both
avian parents and nest contents conspicuous to predators
(Colombelli-Négrel and Kleindorfer 2010). Thus, adult-
directed predation risk near nests clearly accentuates the
tradeoff between parental effort and adult survival, and
indeed reduces parental provisioning rates in a variety of
avian species (Ghalambor and Martin 2000, 2001, Peluc
et  al. 2008, Tilgar et  al. 2011). Further, past studies have
documented various factors that affect parental responses to
predators near avian nests, notably nest concealment (Eggers
et al. 2006, Peluc et al. 2008, Lima 2009). Nest concealment
may be important to understanding how conspicuously
ornamented birds respond to both adult-directed and nest
predation risk near nests, since conspicuous individuals may
be more likely to maintain parental effort under risk if nests
are more concealed.
To assess whether adult-directed predation risk affects
the relationship between sexual ornamentation and parental
effort, we manipulated perceived adult-directed predation
risk near nests, quantified nest concealment, and measured
sexual pigmentation in yellow warblers Setophaga petechia.
The yellow warblers is a biparental passerine with colorful
pigmentation and moderate sexual dichromatism. Males
express more intense carotenoid-based pigmentation and
more phaeomelanin-based breast streaking than females,
although considerable variation in pigmentation occurs
within each sex (Lowther et al. 1999, Grunst et al. 2014a).
The intensity of carotenoid-based pigmentation and the
amount of phaeomelanin-based pigmentation are largely
uncorrelated in males, and positively correlate in females
(r  0.41), but with significant independent variation. Thus,
the two pigment types may convey non-equivalent informa-
tion (Grunst et  al. 2014a). We have previously established
that carotenoid- and phaeomelanin-based pigmentation are
associated with metrics of individual condition in both sexes
of the yellow warbler (Grunst et al. 2014a, b), and warblers
show assortative social pairing by pigmentation (Grunst
and Grunst 2014). Further, a past study found greater extra-
10
pair paternity among males expressing more melanin-based
pigmentation (Yezerinac and Weatherhead 1997), and in
our study population males co-expressing high levels of
both melanin- and carotenoid-based pigmentation lose less
paternity in their social broods (Grunst and Grunst 2014).
Thus, as also reported for both carotenoid- and melanin-
based pigmentation in other species, sexual pigmentation in
the yellow warbler may serve as a sexual signal of individual
condition (Hill 1991, Badyaev and Duckworth 2003,
Siefferman and Hill 2003, Safran and McGraw 2004, Dunn
et al. 2010) or genetic quality (Roulin et al. 2001, Boerner
and Krüger 2009), in addition to making individuals
conspicuous to predators. As a sexual signal, pigmentation
may predict both individual behavioral performance and the
parental allocation decisions of the social mate.
We made specific predictions regarding how adult-
directed predation risk would alter relationships between
sexual pigmentation and the parental behavior of the focal
individual and social mate. First, we predicted that preda-
tion risk might induce a strong positive relationship between
the pigmentation (either carotenoid- or melanin-based) of
the social mate and parental effort, by causing individuals
with low quality, or ‘unattractive’ mates to abandon paren-
tal effort. Further, we reasoned that highly pigmented birds
might have mates more willing to take parental risks (Matessi
et al. 2009), and might also be highly conspicuous to preda-
tors (Cabido et al. 2008, Journey et al. 2013), particularly
if expressing intense carotenoid-based pigmentation, which
covers the majority of the body. Thus, compared to less
pigmented individuals, we predicted that highly pigmented
birds might show lower levels of parental effort under preda-
tion risk, but might show higher parental performance under
normal, low risk conditions. However, we also predicted that
highly pigmented birds might be more likely to maintain
parental effort under predation risk at highly concealed
nests. In addition, birds with more phaeomelanin-based
pigmentation might be more likely to take parental risks,
given a genetic correlation between phaeomelanin-based
pigmentation, aggression and boldness, as suggested by some
previous research (Studd and Robertson 1985a, b, Quesada
and Senar 2007, Ducrest et al. 2008). On the other hand, past
studies in the yellow warbler found that phaoemelanin-based
pigmentation positively correlated with territorial aggression but
negatively correlated with paternal effort, including nest defense
against a snake (Studd and Robertson 1985a, b). This past
research suggests that highly melanic birds might display lower
parental motivation and thus take few parental risks. Finally, we
predicted that females might take more parental risks than males
due to higher certainty of parentage (Yezerinac and Weatherhead
1997), and less intense coloration that would make them less
conspicuous to predators (Martin and Badyaev 1996). There-
fore, greater inter-sex differences in parental effort might occur
under predation risk than under normal conditions.
Methods
Study system and field methods
During the 2010 to 2012 breeding seasons, we studied
yellow warblers breeding along riparian corridors at the
Univ. of California’s Sierra Nevada Aquatic Research
Laboratory (SNARL; 37°36′51′′N, 118°49′47′′W), and
in the adjacent Inyo National Forest. Warblers arrive on
the breeding grounds in May, and nesting occurs from mid-
May to early July. Pairs re-nest following nest depredation
but fledge only one clutch per season. We captured birds in
mist nets using conspecific playback during the pre-nesting
period, or by placing nets near nests, and marked birds with
a United States Geological Survey (USGS) aluminum band
and colored leg bands. We located nests through behavioral
observation and checked nests every 2–3 d to monitor nest
contents. After nests fledged or failed, we scored nest con-
cealment on a scale of 0 to 2. A score of zero indicated that
the nest was 100–80% visible (exposed) when viewed from
2 m in front of the substrate plant, 0.5 indicated 80–60%
exposure, 1, 60–40% exposure, 1.5, 40–20% exposure, and
2, 20–0% exposure.
Measurement of pigmentation
We collected 5 feathers bearing carotenoid pigmentation
from non-adjacent breast regions, and stored feathers in
closed envelopes until spectrometric analysis. We arranged
feathers on a black felt background (with zero reflectance),
and then used an USB4000 spectrometer with a xenon light
source (range: 200–1100 nm; Ocean Optics, Dunedin, FL,
USA) to obtain reflectance spectra between 300–725 nm,
across the avian visual range (Montgomerie 2006, Hegyi
et al. 2007, Grunst et al. 2014a). We characterized reflectance
spectra using colorimetric measurements of total reflectance
(brightness), saturation (chroma or spectral purity), and
hue (spectral location). Carotenoid pigmentation displays a
bimodal reflectance spectrum, with peaks of reflectance for
both ultraviolet and yellow (or red) light, and high absor-
bance of blue–green light. Thus, to characterize reflectance
spectra, we calculated carotenoid saturation (chroma), blue
saturation, ultraviolet saturation, average reflectance, and
lambda 50 (a measurement of hue), as described by Anders-
son and Prager (2006). We performed a principal compo-
nent analysis to derive a single factor (PC1) descriptive of
variance in reflectance spectra (Montgomerie 2006, Hegyi
et  al. 2007, Grunst et  al. 2014a). Loadings on PC1 were
0.51 for carotenoid saturation, 0.48 for ultraviolet satura-
tion, 0.52 for blue saturation, 0.06 for total reflectance,
and 0.47 for lambda 50, with an eigenvalue of 3.61 and
72.5% of total variation explained. Previous studies suggest
that increasing the concentration of yellow carotenoids in
feathers increases lambda 50 (resulting in a shift towards
orange), ultraviolet saturation, and carotenoid saturation,
but reduces average reflectance and blue saturation (Anders-
son and Prager 2006). Thus, PC1 positively correlates with
the concentration of carotenoids in feathers. We corrected
for a year effect on carotenoid PC1. This correction did not
qualitatively alter statistical results.
Further, we used a Stylus 800 Olympus camera to take
digital photographs of the front and sides of all birds. For
males, we used the threshold colour function in ImageJ to
extract the percent coverage of red–brown (phaeomelanin)
from photographs (Schneider et al. 2012, see Grunst et al.
2014a for further detail). For most females, melanin cover-
age was too low to accurately quantify in ImageJ. Thus, we
scored female melanin coverage on a scale of 0 to 2. A score
of zero corresponded to no melanin-based streaking, 0.5
to a trace of streaking, 1 to moderate streaking (∼ 1–2%
coverage), and 2 to heavy streaking (∼ 5% coverage).
Predator presentation experiment
We experimentally increased perceived adult-directed preda-
tion risk at nests by presenting a recording and taxidermic
mount of a sharp-shinned hawk Accipiter striatus. A decoy
and recording of a house finch Haemorhous mexicanus served
as a negative control. We manually compiled hawk and finch
recordings, such that both types of recording consisted of
3 min of vocalization followed by 3 min of silence. We used
the same recording for all experiments, but randomly
used one of three hawk decoys and one of three finch
decoys. We placed decoys ∼ 6–10 m from nests and ∼ 2 m
off the ground (Ghalambor and Martin 2000, 2001, Peluc
et  al. 2008), and projected recordings using MP3 players
placed in the vegetation next to decoys. We also recorded
nests without any decoy present, to assess baseline behavior.
We performed recordings (baseline, hawk, finch) in ran-
domly determined order for sequential periods of ∼ 1.5 h
each with ∼ 10 min between recordings, using Canon 800
series camcorders with 124-min tapes. We erected cam-
corders on tripods concealed ∼ 5 m from nests. We initiated
recordings between 06:00 and 16:00 Pacific Daylight Time
(PDT) due to high recording volume, and controlled for
time of recording in statistical analyses. We performed the
experiment on day 4–7 of the nestling period. Overall, we
performed the experiment at 71 nests of 57 males and 67
females, but performed finch presentations only 45 times,
after it became clear that they did not differ from controls
(Results). Further, sample sizes for analyses (Results) are
reduced in some cases because we did not have pigmenta-
tion measurements from all birds. From video-recordings,
we determined paternal and maternal provisioning rates.
We distinguished the sexes based on plumage differences
and colored leg bands.
Statistical analysis
We performed linear mixed effect models (LMMs) in R
2.15.2 (packages lme4 and lmerTest) (Bates et  al. 2012,
Kuznetsova et  al. 2013), with nest, male, and female
identity as random effects. We sequentially removed non-
significant predictors (a  0.05) from models and obtained
final p-values using type III F-tests with Satterthwaite
approximations for degrees of freedom. First, we constructed
separate LMMs to predict paternal and maternal provision-
ing rate from treatment (baseline, finch, hawk), treatment
order (hawk presented first or subsequently), pigmentation
of the focal bird, brood size, nestling age, nest conceal-
ment, and time of day. We included three-way interactions
between pigmentation variables, nest concealment, and treat-
ment, to assess whether pigmentation and nest concealment
combine to predict the response to predation risk near nests.
We also included a two-way interaction between treat-
ment and treatment order, since treatment order might
affect behavior primarily in the baseline recording (due to
carryover effects from the predator presentation). Second,
11
we used separate models, constructed in the same way, to
assess whether the social mate’s sexual pigmentation inter-
acted with treatment to predict the focal individual’s provi-
sioning effort. Finally, to assess the relationship between sex
and parental risk-taking, we constructed an LMM to predict
parental provisioning rate from the interaction between sex,
treatment, and nest concealment, with the same covariates
included as above (but not sexual pigmentation). We square-
root transformed provisioning rates to normalize model
residuals, and centered continuous predictor variables to
facilitate the interpretation of main effects when including
interactions in the model (Schielzeth 2010).
Results
Pigmentation of the focal individual and parental
risk-taking
Birds did not provision nestlings at different rates during the
negative control (finch) treatment compared to at baseline
(LMM: males: t145  0.12, p  0.89, females: t145  0.17,
p  0.26), so we collapsed baseline and negative control into
a single category.
Males strongly reduced paternal provisioning rate under
adult-directed predation risk and reduced provisioning rate
less under risk at highly concealed nests (Table 1, Fig. 1).
Further, the effect of nest concealment on the response
to the predator was stronger for males with more colorful
carotenoid-based pigmentation than for males with less
intense carotenoid pigmentation (significant treatment 
carotenoid PC1  nest concealment interaction, Table 1).
Indeed, the treatment  nest concealment interaction was
highly significant when restricting the dataset to the most
intensely pigmented half of males (males with pigmenta-
tion above the median value; F1, 58  7.28, b  0.54  0.20,
p  0.009), but was non-significant when restricting
the dataset to the least intensely pigmented half of males
(F1, 58  2.35, b  0.42  0.25, p  0.09). Thus, compared
to less pigmented males, males with more intense carotenoid-
based pigmentation were more likely to maintain paternal
provisioning rates under predation risk at highly concealed
nests, but displayed large reductions in provisioning rate
under risk at highly exposed nests (Fig. 2). Male melanin
Table 1. LMM predicting paternal provisioning rate from male carotenoid-based pigmentation, predation risk treatment, and nest concealment.
Estimate (b  SE)
Intercepta 2.10  0.33
Treatmentb 1.11  0.12
Carotenoid PC1 0.11  0.11
Nest concealment 0.001  0.10
Treatment orderc 0.35  0.15
Time 0.07  0.03
Treatment  carotenoid PC1 0.01  0.11
Treatment  nest concealment 0.33  0.11
Carotenoid PC1  nest concealment 0.12  0.10
Treatment  carotenoid PC1  concealment 0.32  0.12
­n  176 observations, 66 nests, 65 females, 56 males.
ap value from initial LMM output, not an F test.
bPredation risk treatment contrasted to baseline.
cPredation treatment first contrasted to baseline first.
12
pigmentation was not associated with paternal responses
to predation risk near nests (LMM: treatment  melanin,
F1, 110  0.23, b  0.06  0.13, p  0.62), regardless of
nest concealment (treatment  melanin  nest concealment,
F1, 110  0.28, b  0.08  0.12, p  0.50). Further, nestling
age (F1, 131  2.53, b  0.12  0.07, p  0.11) and brood
size (F1, 78  0.23, b  0.03  0.06, p  0.62) were not sig-
nificantly related to paternal provisioning rate (Supplemen-
tary material Appendix 1, Table A1, initial full model), but
the final model retained effects of treatment order and time
of day (Table 1).
Like males, females reduced provisioning rate under
adult-directed predation risk (Table 2, Fig. 1), and reduced
provisioning less under risk at more concealed nests (Table
2). Further, females expressing more melanin-based pig-
mentation reduced provisioning rates under predation
risk less than other females (Table 2, Fig. 3), and this
effect was independent of nest concealment (LMM: treat-
ment  melanin score  nest concealment; F1, 67  1.22,
b  0.16  0.14, p  0.27). Female carotenoid-based
pigmentation was not related to maternal responses to
predation risk near nests (treatment  carotenoid PC1;
F1, 65  2.47, b  –0.24  0.16 p  0.12), irrespective of
nest concealment (treatment  carotenoid PC1  nest
concealment, F1, 64  0.27, b  0.07  0.15, p  0.59).
Nestling age (F1, 98  2.33, b  0.13  0.08, p  0.13), treat-
ment order (F1, 110  0.001, b  0.19  0.19, p  0.98), and
time of day (F1, 98  2.05, b  0.06  0.04, p  0.15) were
also unrelated to maternal provisioning behavior (Supple-
mentary material Appendix 1, Table A2, initial full model),
but the final model retained a positive effect of brood size on
maternal provisioning rate (Table 2).
Pigmentation of the social mate and parental
risk-taking
When faced with adult-directed predation risk, warblers
with more intensely pigmented social mates did not reduce
parental effort less than birds with less pigmented mates.
Males did not maintain higher provisioning effort under
risk when paired to females with more intense carotenoid
pigmentation (LMM: treatment  female carotenoid PC1;
F1, 63  0.04, b  0.03  0.17, p  0.83;), or more mela-
nin pigmentation (treatment  female melanin score;
F DF (Denom.) p( F)
– –  0.001
85.49 117.86  0.001
1.14 75.87 0.28
2.96 95.29 0.08
5.04 167.86 0.02
5.74 156.75 0.01
0.01 117.14 0.91
7.94 117.10 0.005
0.11 155.24 0.73
6.20 117.94 0.01
 Table 2. LMM predicting maternal provisioning rate from female
melanin-based pigmentation, predation risk treatment, and paternal
provisioning rate.

Estimate DF
(b  SE) F (Denom.) p( F)
Intercepta 2.24  0.25 – –  0.001
Treatmentb 0.88  0.13 45.41 72.02  0.001
Melanin score 0.08  0.13 2.44 38.55 0.12
Nest concealment 0.04  0.12 0.37 40.30 0.54
Brood size 0.22  0.08 7.54 35.17 0.009
Treatment  melanin 0.32  0.13 13.05 72.51  0.001
score
Treatment   0.47  0.13 5.44 72.59 0.02
concealment

n  120 observations, 45 nests, 42 females, 40 males.

ap value from initial LMM output, not an F test.
bCoefficient estimate for predation risk treatment contrasted to

Figure 1. Adult-directed predation risk reduced nestling provision- baseline.

ing rates in both sexes. Means are back transformed and include
only replicates for which finch (negative control) presentations were
performed (n  45). Error bars represent 95% confidence intervals.
F1, 65  2.00, b  0.25  0.18, p  0.16), and female pig-
mentation did not correlate with paternal provisioning rate
across treatments (non-significant main effects; p  0.30;
Supplementary material Appendix 1, Table A3, full model).
Similarly, male carotenoid pigmentation (treatment  male
carotenoid PC1; F1, 105  1.23, b  0.14  0.12, p  0.26)
and melanin pigmentation (treatment  male melanin
coverage; F1, 104  0.08, b  0.03  0.13, p  0.77) were
unrelated to maternal responses to predation risk (Supple-
mentary material Appendix 1, Table A4, full model). How-
ever, across treatments, females provisioned nestlings at higher
rates when paired to males with more intense carotenoid pig-
mentation (LMM: F1, 51  4.13, b   0.19  0.09, p  0.04).
Sex and parental risk-taking
Finally, sex was not related to how warblers reduced
provisioning effort in response to predation risk (LMM:
sex  treatment; F1, 295  1.03, b  0.19  0.19, p  0.30),
irrespective of nest concealment (sex  treatment  nest
concealment; F1, 295  0.02, b  0.03  0.19, p  0.87).
Further, males and females did not display different levels of
provisioning effort across treatments, as indicated by a non-
significant main effect of sex (F1, 295  0.37, b  0.03  0.11,
p  0.54, Fig. 1; Supplementary material Appendix 1,
Table A5, full model).
Discussion
Our study demonstrates that sexual pigmentation in both
males and females may be associated with changes in paren-
tal provisioning rates in response to adult-directed predation
risk near nests. Further, for males, our results suggest that
nest concealment modifies the relationship between con-
spicuous carotenoid-based pigmentation and maintenance
of parental effort under predation risk. Past researchers have
suggested a genetic correlation between melanin-based pig-
mentation, stress resistance, and the boldness of parents

Figure 2. Relationship between male carotenoid-based pigmentation and paternal provisioning rate (feeding trips h1, square-root trans-
formed) under predation risk versus baseline conditions at nests with low concealment (a), versus high concealment (b). Residual provision
rate (y-axis) controls for treatment order, time of day, and variation in nest concealment within the low and high concealment categories.
Shaded regions within dotted lines represent 95% confidence intervals.

13

Figure 3. Relationship between female melanin score and maternal
provisioning rate under predation risk versus baseline conditions.
Residual provisioning rate (Square root h1) controls for brood
size and variation in nest concealment within the low and high
nest concealment categories. Shaded regions within dotted lines
represent 95% confidence intervals.
when defending nests against predators (Quesada and Senar
2007, Ducrest et  al. 2008, Boerner and Krüger 2009, van
den Brink et  al. 2012). Further, in non-parental contexts,
researchers have found that individuals with more intense
carotenoid-based pigmentation are bolder under predation
risk (Godin and Dugatkin 1996), perhaps because they
better evade predation (Cabido et al. 2008). However, con-
spicuous coloration might also increase the risk of detec-
tion by predators, and thus induce stronger responses to
predation threats, especially where vegetative cover is lim-
ited. Indeed, a recent study documented that species with
more conspicuous coloration (origin of coloration was not
considered) respond more intensely to the vocal calls of a
raptor predator (the broad-winged hawk Buteo platypterus,
Journey et  al. 2013). However, few studies document a
relationship between sexual pigmentation and changes in
parental provisioning behavior under adult-directed pre-
dation risk within a species, or indicate that an interaction
between nest concealment and sexual pigmentation is associ-
ated with parental responses to adult-directed predation risk.
First, our results suggest that males with colorful caro-
tenoid-based pigmentation respond to adult-directed pre-
dation risk near nests in a fashion that depends on nest
concealment. Compared to less intensely pigmented males,
males with more intense carotenoid-based pigmentation
displayed greater maintenance of nestling provisioning rates
under predation risk at concealed nests, but showed greater
reductions in paternal effort under risk at highly exposed
nests (Fig. 2). These results suggest that males with intense
carotenoid pigmentation are ‘good parents’ under predation
risk at concealed nests, but not exposed nests. We predicted
that highly pigmented males might display large reductions
in nestling provisioning rate under predation risk because
these birds (especially males) might be more conspicuous to
predators (Martin and Badyaev 1996), and this appeared to
occur at highly exposed nests. However, we also predicted
that highly pigmented males might be more likely to main-
tain provisioning effort at concealed nests, in accordance
with results. In our population, carotenoid pigmentation
14
increases longitudinally with male age (Grunst et al. 2014a,
b), and is greater in first year breeding males that are in
better condition as nestlings and at pre-breeding molt
(Grunst et  al. 2014a). Thus, carotenoid-based pigmenta-
tion may be related to quality and age-related differences
in both the ability to evade predation, and experience with
assessing predation threats (Cabido et al. 2008, Fowler-Finn
and Hebets 2011). Therefore, at concealed nests, males with
intense carotenoid pigmentation may maintain higher pro-
visioning effort under predation risk than less pigmented
males due to a better capacity to evade predation combined
with more experience with assessing how predation risk
varies with vegetation cover. In contrast, at highly exposed
nests, males with intense carotenoid pigmentation may
show large reductions in paternal effort under predation risk
because the risk of detection and depredation is too high.
Nest concealment and predation risk in combination may
thus affect the paternal benefits that females obtain from
conspicuously pigmented males.
Notably, carotenoid-based pigmentation in yellow
warblers appears more conspicuous than melanin-based
pigmentation, which is restricted to narrow streaks on the
breast. Thus, one might expect carotenoid rather than mela-
nin pigmentation to be associated with the relationship
between nest concealment and paternal responses to preda-
tion risk near nests, in accordance with our results. However,
we did not actually quantify how carotenoid-based pigmen-
tation affects the conspicuousness of birds, which would
require quantifying the contrast between background color-
ation and birds differing in carotenoid-based pigmentation
(Delhey et al. 2010). Thus, our results could reflect differ-
ences in the capacity of males to assess how nest conceal-
ment affects the predation threat, independent of differences
in conspicuousness between individuals. Future research to
clarify this point would be interesting.
In contrast to males, in females melanin-based pigmen-
tation was related to responses to predation risk, but this
relationship was independent of nest concealment. Results
support the hypothesis that melanin-based pigmentation
may indicate a willingness to take parental risks (Quesada and
Senar 2007, Da Silva et al. 2013), since females with more
melanin pigmentation reduced maternal provisioning rate
under risk less than other females. Past research has linked
melanin-based pigmentation with greater boldness and risk-
taking, and researchers have suggested a genetic correlation
between the two (Quesada and Senar 2007, Ducrest et  al.
2008, Mafli et al. 2011, Da Silva et al. 2013). Indeed, a cross
fostering experiment in barn owl Tyto alba nestlings showed
that antipredatory behavior and melanin-based pigmenta-
tion are correlated and heritable (van den Brink et al. 2012).
Thus, although the work in barn owls involved a eumelanin-
based (black) trait, phaeomelanin-based pigmentation in
female yellow warblers may also reflect genetic differences
in antipredator strategy and risk responsiveness. However,
it is unclear why male melanin-based pigmentation did not
also correlate with paternal responses to the predation threat,
given a genetic association between melanin pigmentation
and risk-taking behavior.
Alternatively, females with more melanin-based pigmen-
tation might reduce provisioning under predation risk less
than other females because melanin-based pigmentation
is related to individual quality, a better capacity to evade
predation, and greater parental motivation. Indeed, in our
population, females with more phaeomelanin pigmenta-
tion are in better condition at pre-breeding molt and are
less likely to be first year breeders (Grunst et al. 2014a),
and phaeomelanin pigmentation has also been associ-
ated with metrics of individual quality in species such as
barn swallows Hirundo rustica (Safran and McGraw 2004,
Galván and Møller 2013, but see Saino et  al. 2013) and
barn owls Tyto alba (Roulin et  al. 2008). Further, past
studies in eastern bluebirds Sialia sialia and barn owls
have associated phaeomelanin-based pigmentation with
greater parental effort (Roulin et al. 2001, Siefferman and
Hill 2003), although past research in male yellow warblers
found a contrary result (Studd and Robertson 1985a, b,
but see Lozano and Lemon 1996).
Our results were contrary to the prediction that predation
risk might accentuate differential allocation (Matessi et  al.
2009), since neither males nor females showed greater main-
tenance of parental effort under predation risk when paired
to highly pigmented mates. Increasing parental risk-taking
when paired to a highly ornamented mate could be adaptive
if ornamentation positively correlates with offspring genetic
or phenotypic quality (Burley 1986), or if individuals risk
abandonment by ‘attractive’ partners by not taking paren-
tal risk (Matessi et  al. 2009). However, our results suggest
that differential allocation is not particularly strong under
increased risk to the adult (Mahr et al. 2012), which might
indeed be expected if acute risks to survival lower variance
in parental effort. On the other hand, we found some sup-
port for the idea that birds adjust levels of parental effort to
the ornamentation of their mates, as predicted by the dif-
ferential allocation hypothesis (Burley 1986, Ratikainen and
Kokko 2010). Specifically, females paired to males with more
intense carotenoid-based pigmentation provisioned nestlings
at higher rates across treatments. However, as a caveat, this
result might also reflect superior territory or mate quality
among males with more intense carotenoid pigmentation
(Sheldon 2000), particularly since warblers assortatively pair
by both carotenoid- and melanin-based pigmentation in our
study population (Grunst and Grunst 2014).
Our results also did not suggest differential parental risk-
taking by males and females (Michl et al. 2000). Males and
females did not take different levels of parental risk even
at highly exposed nests, and also did not provide different
levels of parental care at baseline. We predicted that males
might provide less parental care than females under preda-
tion risk. First, under a predation threat, males might take
less parental risk than females because more colorful sexual
pigmentation makes them more conspicuous and vulner-
able to predation (Zuk and Kulluru 1998, Stuart-Fox et al.
2003, Journey et al. 2013, but see Bókony et al. 2008). It is
unclear why greater conspicuousness of males did not lead
to lower provisioning rates under predation risk in compari-
son to females, especially at highly exposed nests. However,
this result might arise due to variance in coloration within
the sexes (Grunst et  al. 2014a), because males risk losing
social mates by not demonstrating comparable willingness
to take parental risks (Matessi et  al. 2009), or if males are
overall less risk sensitive than females (Breitwisch and Hudak
1989, Ensminger and Westneat 2012). Further, males might
also respond more to predators near nests than females if
males pursue a reproductive strategy focused on multiple
mating rather than parental care (Kokko 1998, Magrath
and Komdeur 2003), and have low certainty of parentage
(Yezerinac et  al. 1996, Yezerinac and Weatherhead 1997).
The high rate of extra-pair paternity in our study popula-
tion renders this later hypothesis particularly feasible (30.7%
of nestlings are extra-pair offspring, Grunst and Grunst
2014). However, previous studies have also found that
male yellow warblers do not provide less parental care than
females, despite high rates of extra-pair paternity, potentially
due to relatively low between-season survival rates (Yezerinac
et  al. 1996, Lozano and Lemon 1996). We hypothesized
that higher costs of parental care under predation risk could
induce parental differences between the sexes. However,
equal parental effort under baseline conditions suggests that
the two sexes have equivalent investments in the social brood.
Males and females might thus weight costs to survival against
parental care in a similar fashion under predation risk.
Finally, our study involved parental risk-taking under
an adult-directed predation risk, whereas others have also
considered responses to nest predation risk (Ghalambor
and Martin 2000, Peluc et al. 2008). Given nest predation
risk, parents reduce provisioning rate at exposed nests
more than at concealed nests, but to prevent nest detection
(Ghalambor and Martin 2000), rather than depredation of
the adult. Thus, the interaction between male carotenoid-
based pigmentation, nest concealment, and nest visitation
rate might have been similar if we had presented a predator
that only targets nests, instead of the sharp-shinned hawk.
On the other hand, the relationship between melanin-based
pigmentation and maternal provisioning rate under adult-
directed predation risk might arise because highly melanic
females are less sensitivity to highly stressful situations
(Almasi et  al. 2008, Ducrest et  al. 2008). Thus, this rela-
tionship might not be present under a nest predation threat,
which does not pose a direct threat to the adult and might
thus be less stressful.
In summary, our results suggest that both carotenoid-
and melanin-based pigmentation may be associated with
parental responses to adult-directed predation risk near
nests. The relationship between male carotenoid-based
pigmentation and paternal provisioning under predation
risk depended on nest concealment, with highly pigmented
males maintaining higher levels of nestling provisioning
under predation risk only at highly concealed nests. Thus,
our results suggest that nest concealment and predation risk
interact to determine the paternal benefits that females receive
from males with intense carotenoid-based pigmentation. In
contrast, females with more melanin-based pigmentation
showed greater maintenance of provisioning effort under
predation risk irrespective of nest concealment, suggesting
that female melanin-based pigmentation is associated with a
willingness to take parental risks. Finally, neither males nor
females maintained higher parental effort under predation
risk when paired to a highly pigmented mate, despite an
overall positive association between maternal provisioning
rate and male carotenoid pigmentation. Therefore, in this
species, differential allocation does not appear to be accentu-
ated by predation risk, but the sexual pigmentation of focal
individuals is associated with parental risk-taking.
15
­­­­­
Acknowledgements – We thank the Sierra Nevada Aquatic Research
Laboratory (SNARL) and the Inyo National Forest for providing
facilities and access to field sites, with particular thanks to SNARL’s
director, Daniel Dawson. Thanks also to Stephen Myers, who
sponsored the bird-banding sub-permit (United States Geological
Survey) essential to this research, and to Marlene Zuk, Daphne
Fairbairn, and Derek Roff for comments that aided in improving
the manuscript. Susan Grunst provided invaluable assistance in the
field, and undergraduate assistants from the Univ. of California,
Riverside aided in data processing. The Univ. of California Natural
Reserve System’s Mildred E. Mathias research grant, Sigma Xi,
and the Valentine Eastern Sierra Reserve’s graduate student
research grant provided funding. The Univ. of California,
Riverside’s Animal Care and Use Committee approved all protocols
(Protocol A-20100003E). Fieldwork was further authorized by a
USGS bird banding permit (23035-G), a California state collecting
permit (SC11060), a federal migratory bird collecting permit
(MB22669A-0), and a special use permit through the Inyo National
Forest (MLD100007P).
References
Almasi, B., Roulin, A., Jenni-Eiermann, S. and Jenni, L. 2008.
Parental investment and its sensitivity to corticosterone is
linked to melanin-based coloration in barn owls. – Horm.
Behav. 54: 217–223.
Andersson, M. and Prager, M. 2006. Quantifying color. – In:
Hill, G. and McGraw, K. (eds), Bird coloration. Vol. 1.
Mechanisms and measurements. Harvard Univ. Press,
pp. 90–147.
Badyaev, A. and Duckworth, R. 2003. Context-dependent sexual
advertisement: plasticity in the development of sexual
ornamentation throughout the lifetime of a passerine bird.
– J. Evol. Biol. 16: 1065–1076.
Bates, D., Maechler, M. and Bolker, B. 2012. lme4: linear mixed-
effects models using S4 classes. – R package ver. 0.999999-0,
 http://CRAN.R-project.org/package  lme4.
Boerner, M. and Krüger, O. 2009. Aggression and fitness differ-
ences between plumage morphs in the common buzzard (Buteo
buteo). – Behav. Ecol. 20: 180–185.
Bókony, V., Liker, A., Lendvai, Á. Z. and Kulesár, A. 2008. Risk-
taking and survival in the house sparrow Passer domesticus: are
plumage ornaments costly? – Ibis 150: 139–151.
Breitwisch, R. and Hudak, J. 1989. Sex differences in risk-taking
behavior in foraging flocks of house sparrows. – Auk 106:
150–153.
Burley, N. 1986 Sexual selection for aesthetic traits in species with
biparental care. – Am. Nat. 127: 415–445.
Cabido, C., Galan, P., Lopez, P. and Martin, J. 2008. Conspicu-
ousness-dependent antipredatory behaviour may counteract
coloration differences in Iberian rock lizards. – Behav. Ecol.
20: 362–370.
Candolin, U. 1998. Reproduction under predation risk and the
tradeoff between current and future reproduction in the stick-
leback. – Proc. R. Soc. B 265: 1171–1175.
Colombelli-Négrel, D. and Kleindorfer, S. 2010. Video nest mon-
itoring reveals male coloration-dependant nest predation and
sex differences in prey size delivery in a bird under high sexual
selection. – J. Ornithol. 151: 507–512.
Da Silva, A., Van den Brink, V., Emaresi, G., Luzio, E., Bize, P.,
Dreiss, A. N. and Roulin, A. 2013. Melanin-based colour
polymorphism signals agressive personality in nest and terri-
tory defence in the tawny owl (Strix aluco). – Behav. Ecol.
Sociobiol. 67: 1041–1052.
Delhey, K., Roberts, M. L. and Peters, A. 2010. The carotenoid-
continuum: carotenoid-based plumage ranges from conspicu-
ous to cryptic and back again. – BMC Ecol. 10: 13.
16
Ducrest, A.-L., Keller, L. and Roulin, A. 2008. Pleiotropy in the
melanocortin system, coloration, and behavioral syndromes.
– Trends Ecol. Evol. 29: 502–510.
Dunn, P. O., Garvin, J. C., Whittingham, L. A., Freeman-Gallant,
C. R. and Hasselquist, D. 2010. Carotenoid and melanin-
based ornaments signal similar aspects of male quality in
two populations of common yellowthroat. – Funct. Ecol. 24:
149–158.
Eggers, S., Griesser, M., Nystrand, M. and Ekman, J. 2006. Preda-
tion risk induces changes in nest-site selection and clutch size
in the Siberian jay. – Proc. R. Soc. B 273: 701–706.
Ensminger, A. L. and Westneat, D. F. 2012. Individual and sex
differences in habituation and neophobia in house sparrows
(Passer domesticus). – Ethology 118: 1085–1095.
Fowler-Finn, K. and Hebets, E. 2011. More ornamented males
exhibit increased predation risk and antipredatory escapes, but
not greater mortality. – Ethology 117: 102–114.
Galván, I. and Møller, A. P. 2013. Phaeomelanin-based coloration
predicts survival rates in birds. – Physiol. Biochem. Zool. 86:
184–192.
Ghalambor, C. and Martin, T. 2000. Parental investment
strategies in two species of nuthatch vary with stage specific
predation risk and reproductive effort. – Anim. Behav. 60:
263–267.
Ghalambor, C. and Martin, T. 2001. Fecundity-survival tradeoffs
and parental risk-taking in birds. – Science 20: 494–497.
Godin, J.-G. J. and Dugatkin, L. A. 1996. Female mating prefer-
ence for bold males in the guppy, Poecilia reticulate. – Proc.
Natl Acad. Sci. USA 93: 10262–10267.
Götmark, F. and Olsson, J. 1997. Artificial colour mutation: do
red-painted great tits experience increased or decreased preda-
tion? – Anim. Behav. 53: 83–91.
Grunst, A. S. and Grunst, M. L. 2014. Multiple sexual pigments,
assortative social pairing, and genetic paternity in the yellow
warbler (Setophaga petechia). – Behav. Ecol. Sociobiol. 68:
1451–1463.
Grunst, A. S., Rotenberry, J. T. and Grunst, M. L. 2014a. Age-
dependent relationships between multiple sexual pigments and
condition in males and females. – Behav. Ecol. doi: 10.1093/
beheco/art124
Grunst, A. S., Salgado-Ortiz, J., Rotenberry, J. T. and Grunst, M.
L. 2014b. Phaeomelanin- and carotenoid-based sexual pig-
mentation reflect oxidative status in two populations of the
yellow warbler (Setophaga petechia). – Behav. Ecol. Sociobiol.
doi: 10.1007/s00265-013-1681-8
Harris, W. and Uller, T. 2009. Reproductive investment when mate
quality varies: differential allocation versus reproductive com-
pensation. – Phil. Trans. R. Soc. B 364: 1039–1048.
Hegyi, G., Szigeti, B., Török, J. and Eens, M. 2007. Melanin,
carotenoid, and structural plumage ornaments: information
content and role in great tits Parus major. – J. Avian Biol. 38:
698–708.
Hill, G. E. 1991. Plumage color is a sexually selected indicator of
male quality. – Nature 350: 337–339.
Hoelzer, G. A. 1989. The good parent process of sexual selection.
– Anim. Behav. 38: 1067–1078.
Journey, L., Drury, J. P., Haymer, M., Rose, K. and Blumstein, D.
T. 2013. Vivid birds respond more to acoustic signals of pred-
ators. – Behav. Ecol. Sociobiol. 67: 1285–1293.
Kokko, H. 1998. Should advertising parental care be honest?
– Proc. R. Soc. B 265: 1871–1878.
Kuznetsova, A., Brockhoff, P. B. and Christensen, R. H. B. 2013.
lmerTest: tests for random and fixed effects for linear mixed
effect models (lmer objects of lme4 package). – R package ver.
1.1-0,  http://CRAN.R-project.org/package  lmerTest.
Lima, S. L. 2009. Predators and the breeding bird: behavioral and
reproductive flexibility under the risk of predation. – Biol. Rev.
84: 485–513.
Lowther, P., Celada, C., Klein, N., Rimmer, C. and Spector, D.
1999. Yellow warbler (Dendroica petechia). – In: Poole, A.
(ed.), The birds of North America online. Cornell Laboratory
of Ornithology, no. 454.
Lozano, G. A. and Lemon, R. E. 1996. Male plumage, paternal
care, and reproductive success in the yellow warbler Dendroica
petechia. – Anim. Behav. 51: 265–272.
Mafli, A., Wakamatsu, K. and Roulin, A. 2011. Melanin-based
coloration predicts aggressiveness and boldness in captive
Hermann’s tortoise. – Anim. Behav. 81: 859–863.
Magrath, M. and Komdeur, J. 2003. Is male care compromised
by additional mating opportunity? – Trends Ecol. Evol. 18:
424–430.
Mahr, K., Griggio, M., Granatiero, M. and Hoi, N. 2012. Female
attractiveness affects paternal investment: experimental
evidence for male differential allocation in blue tits. – Front.
Zool. 9: 14.
Martin, T. and Badyaev, A. V. 1996. Sexual dichromatism in birds:
importance of nest predation and nest location for females
versus males. – Evolution 50: 2454–2460.
Matessi, G., Caramagnani, C., Griggio, M. and Pilastro, A. 2009.
Male rock sparrows differentially allocate nest defense but not
food provisioning to offspring. – Behaviour 146: 209–223.
Michl, G., Török, J., Garamszegi, L. and Toth, L. 2000. Sex-
dependent risk taking in the collared flycatcher, Ficedula
albicollis, when exposed to a predator at the nestling stage.
– Anim. Behav. 59: 623–628.
Møller, A. and Thornhill, R. 1998. Male parental care, differential
allocation by females and sexual selection. – Anim. Behav. 55:
1507–1515.
Montgomerie, R. 2006. Analyzing colors. – In: Hill, G. and
McGraw, K. (eds), Bird coloration. Vol. 1. Mechanisms and
measurements. Harvard Univ. Press, pp. 90–147.
Peluc, S., Sillett, T., Rotenberry, J. T. and Ghalambor, C. 2008.
Adaptive phenotypic plasticity in an island songbird exposed
to a novel predation risk. – Behav. Ecol. 19: 830–835.
Quesada, J. and Senar, J. C. 2007. The role of melanin and
carotenoid-based plumage coloration in nest defense in the
great tit. – Ethology 113: 640–647.
Ratikainen, I. I. and Kokko, H. 2010. Differential allocation and
compensation: who deserves the silver spoon? – Behav. Ecol.
21: 195–200.
Roulin, A. 2004. The evolution, maintenance and adaptive
function of genetic colour polymorphism in birds. – Biol. Rev.
79: 815–848.
Roulin, A., Dijkstra, C., Riols, C. and Ducrest, A.-L. 2001. Female-
and male-specific signals of quality in the barn owl. – J. Evol.
Biol. 14: 255–266.
Supplementary material (Appendix JAV-00496 at www.
avianbiology.org/readers/appendix). Appendix 1.
Roulin, A., Almasi, B., Rossi-Pedruzzi, A., Ducrest, A.-L.,
Wakamatsu, K., Miksik, I., Blount, J., Jenni-Eiermann, S. and
Jenni, L. 2008. Corticosterone mediates the condition-
dependent component of melanin-based coloration. – Anim.
Behav. 75: 1351–1358.
Safran, R. and McGraw, K. 2004. Plumage coloration, not length
or symmetry of tail-streamers, is a sexually selected trait in
North American barn swallows. – Behav. Ecol. 15: 455–461.
Saino, N., Romano, M., Rubolini, D., Ambrosini, R., Caprioli,
M., Milzani, A., Costanzo, A., Colombo, G., Canova, L. and
Wakamatsu, K. 2013. Viability is associated with melanin-
based color in the barn swallow (Hirundo rustica). – PLoS One
8: e60426.
Schielzeth, H. 2010. Simple means to improve the interpretability
of regression coefficients. – Methods Ecol. Evol. 1: 103–113.
Schneider, C., Rasband, W. and Eliceiri, K. 2012. NIH image
to ImageJ: 25 years of image analysis. – Nat. Methods 9:
671–675.
Sheldon, B. C. 2000. Differential allocation: tests, mechanisms,
and implications. – Trends Ecol. Evol. 15: 397–402.
Siefferman, L. and Hill, G. 2003. Structural and melanin pigmen-
tation indicate male parental effort and reproductive success
in eastern bluebirds. – Behav. Ecol. 14: 855–861.
Stuart-Fox, D. M., Moussalli, A., Marshal, N. J. and Owens, I. P.
F. 2003. Conspicuous males suffer higher predation risk: visual
modeling and experimental evidence from lizards. – Anim.
Behav. 66: 541–550.
Studd, M. V. and Robertson, R. J. 1985a. Evidence for reliable
badges of status in territorial yellow warblers (Dendroica
petechia). – Anim. Behav. 33: 1102–1113.
Studd, M. V. and Robertson, R. J. 1985b. Sexual selection and
variation in reproductive strategy in male yellow warblers
(Dendroica petechia). – Behav. Ecol. Sociobiol. 17: 101–109.
Tilgar, V., Moks, K. and Saag, P. 2011. Predator-induced stress
changes parental behaviour in pied flycatchers. – Behav. Ecol.
22: 23–28.
van den Brink, V., Dolivo, V., Falourd, X., Dreiss, A. N. and
Roulin, A. 2012. Melanic color-dependent antipredator
behavior strategies in barn owl nestlings. – Behav. Ecol. 23:
473–480.
Yezerinac, S. M. and Weatherhead, P. J. 1997 Extra-pair mating,
male plumage coloration and sexual selection in yellow war-
blers (Dendroica petechia). – Proc. R. Soc. B 264: 527–532.
Yezerinac, S. M., Weatherhead, P. J. and Boag, P. 1996. Cuckoldry
and lack of parentage-dependent paternal care in yellow war-
blers: a cost–benefit approach. – Anim. Behav. 52: 821–832.
Zuk, M. and Kulluru, G. R. 1998. Exploitation of sexual signals
by predators and parasitoids. – Q. Rev. Biol. 73: 415–438.
17